The discovery of RNA interference (RNAi) and the major advances in the understanding of small RNA biology in the past decade have provided researchers with an invaluable tool for wide-scale and rapid genetic screening. As a research tool, RNAi takes advantage of endogenous RNA processing machinery, which permits the silencing of mRNA transcripts with complementary double stranded RNA (dsRNA). This is achieved by introduction of a so-called short interfering (si)RNA complementary to the target gene messenger RNA. Large-scale application of this technology using the robotic platforms developed for small molecule screens has led to the implementation of the first genome-wide forward genetic screens in mammalian cells. We intend to use RNAi screening technology to interrogate the mechanistic basis of immune cell responses to pathogenic insult. Our current efforts are focused on cells of the innate immune system as they form the first line of defense against numerous bacterial and viral pathogens and characterization of these initial encounters are central to our efforts to generate quantitative models of host-pathogen interactions. We are developing assays in macrophage/monocyte cell lines with both microscopy-based 'high content'single cell readouts, and also bioluminescence-based population assays using luciferase reporters driven by inflammatory gene promoters. We have designed and constructed dual promoter lentiviral vectors that permit the expression of two genes from a single virus. This has led to the creation of a RAW264.7 (RAW) mouse macrophage cell line expressing two fluorescent biosensors for high content screening. The first readout expresses a GFP fusion with the RelA NFkB transcription factor driven by its endogenous promoter. This protein partitions to the cytoplasm in unstimulated cells, and translocates to the nucleus in response to a wide range of ligands that promote macrophage activation. Use of the endogenous RelA promoter facilitates a more accurate reproduction of the oscillatory cytosol/nuclear translocation previously observed with endogenous RelA. The second biosensor uses the murine TNF alpha promoter to drive expression of the red fluorescent protein mCherry, fused to a destabilizing PEST sequence which reduces the protein's half-life in the cell. This latter modification provides a more dynamic readout of TNF alpha promoter activity in kinetic experiments. We have established a clonal cell line exhibiting robust RelA translocation in the first hour, followed by a significant increase in TNF alpha promoter-driven mCherry expression after 12 hours in response to a wide range of Toll-like receptor (TLR) ligands. We have use a similar strategy to create a THP1 human monocyte cell line expressing firefly luciferase driven by the human TNF alpha promoter and also renilla luciferase driven by the ubiquitin promoter. The constitutive renilla expression provides a valuable normalization factor for cell number variability in a population-based readout. Thus, the firefly/renilla ratio in this cell line after TLR stimulation provides a measure of TNF alpha promoter activity, and we see a significant and reproducible increase in response to various TLR ligands. Effective delivery of siRNA into hematopoietic cells remains a significant obstacle to the implementation of effective siRNA screens. It is important to establish a reproducible method that can achieve >80% knockdown (KD) of target genes in order to avoid a high frequency of false negatives in genome-wide screens. We have taken advantage of the fact that the creation of the RAW and THP1 reporter cell lines described above provide convenient control siRNA targets in the reporters they express. The ability to assay GFP and renilla luciferase KD in the RAW and THP1 lines respectively provides two advantages. First, it allows us to assay for protein rather than mRNA KD (the most common validation method for siRNA delivery), providing a more direct measure of the required endpoint needed for an effective screen. Second, the ability to run both assays in 384-well format allows for a more extensive matrix of experimental conditions which improves the chance of identifying an optimal delivery protocol. Using such an approach, we have developed highly efficient lipid-based transfection protocols in 384-well format for both of the described RAW and THP1 cell lines. We have also confirmed our assays conform to the plate uniformity criteria established by the NCGC small molecule screening group, and we have identified reproducible positive siRNA controls for a range of TLR ligands, including the TLR receptors, receptor-associated adapter proteins and protein kinases involved in the activation of the MAPK and NFkB signaling pathways. Selected controls of varying phenotypic strength are included on every screening plate, to evaluate the quality metrics as the screens progress. Using the described assay platform, this year we have completed the primary phase of two genome wide screens to identify genes involved in the human and mouse LPS response. As the best characterized TLR stimulus, data from these screens will provide a valuable comparison of the endotoxin response in mouse and human cells, with important potential clinical relevance in the context of septic shock and endotoxin tolerance. For the human THP1 screen, we collected four replicate data points for each gene specific siRNA pool, which required 55 x 384-well plates to complete the screen. siRNA transfection efficiency was evaluated by renilla knockdown on every plate, and averaged >95%. We ran a series of statistical tests to evaluate assay reproducibility. An initial triplicate screen using kinome siRNAs shown a correlation coefficient of R >0.6 for all pairwise replicate plate comparisons. Average Z factors, which should be between 0 and 0.5 for strong controls in an siRNA screen, ranged from 0.2 to 0.35 for the proximal signaling components TLR4, MyD88 and Mal/Tirap. Initial analysis of scores from the genome wide screen show that of the known signaling components involved in the initial response to LPS, the strongest negative scores are observed for the TLR4 receptor complex (TLR4, MD2 and CD14) and the proximal signaling adapter proteins MyD88 and Mal/Tirap. We also observe significant negative scores for the components of the TNF alpha transcriptional enhanceosome (SP1, Egr1 and 2, Jun, ATF and Elk). These data validate our screening protocol and increase the likely physiological significance of novel hits identified in our assay. For the mouse RAW cell screen, we have also recently completed the primary phase of the screen. Data was collected from the initial p65/RelA activation reporter at 40 min, and from the TNF alpha transcriptional reporter at 16 hr. Replicate plates were run for each readout on separate days, and data was considered acceptable if plates showed a spearman rank correlation of R >0.5. The average R value across the screen was 0.71 for the p65/RelA readout and 0.72 for the TNF alpha transcriptional readout. Informatic analysis is underway to assess phenotypes of known TLR4 pathway components and gene lists for secondary screens in both THP1 and RAW cells are being finalized.